MATLAB: Does the documentation for the XCOV function in the Signal Processing Toolbox 6.8 (R2007b) refer to Gaussian random variables as “Uniform noise with mean = 0.5”

Signal Processing Toolbox

The documentation page for the XCOV function contains the following line of code for the example at the bottom of the page:

ww = randn(1000,1);    % Uniform noise with mean = 0.5

The RANDN function is not uniform and has a mean of zero.

Best Answer

This bug has been fixed in Release 2008b (R2008b).

This is an error within the documentation for the XCOV function within the Signal Processing Toolbox for previous releases. The documentation should read as follows:

ww = randn(1000,1);    % White Gaussian noise
[cov_ww,lags] = xcov(ww,10,'coeff');
stem(lags,cov_ww)

Related Solutions

MATLAB: Can the XCOV function return a result that is both unbiased and normalized

The ability to create estimates of the cross-covariance function that are both unbiased and normalized is not available in the Signal Processing Toolbox.

To work around this issue, create unbiased estimates and normalize by the zero lag term manually, as in the following example:

ww = randn(1000,1);
N = 10;
[cov_ww,lags] = xcov(ww,N,'unbiased');
cov_ww = cov_ww / cov_ww(N+1);
stem(lags,cov_ww)

MATLAB: Alternatives to xcov for Image analysis

You can do it yourself, manually. Just write the code - it's not that hard. You can use conv() which is in base MATLAB.

Best Answer

Related Solutions

MATLAB: Can the XCOV function return a result that is both unbiased and normalized

MATLAB: Alternatives to xcov for Image analysis

Related Question